You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hawq.apache.org by "ASF GitHub Bot (JIRA)" <ji...@apache.org> on 2016/01/27 09:34:39 UTC
[jira] [Commented] (HAWQ-366) Analyse data locality ratio for TPCH
query in SFO.
[ https://issues.apache.org/jira/browse/HAWQ-366?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15118863#comment-15118863 ]
ASF GitHub Bot commented on HAWQ-366:
-------------------------------------
GitHub user zhangh43 opened a pull request:
https://github.com/apache/incubator-hawq/pull/300
HAWQ-366. Analyse data locality ratio for TPCH query in SFO.
Data locality ratio for TPCH query(such as Q2) in SFO is low when some machines are down.
The reason for this is that we change net_disk_ratio to 1.0 which leads to we prefer the virtual segment with minimum allocated size plus current split size, but not LOCAL!
To prevent this happens we need to enlarge net_disk_ratio to 1.01. Then if the current size of two vsegs are almost the same, we can choose the LOCAL one.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/zhangh43/incubator-hawq hawq366
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-hawq/pull/300.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #300
----
commit 898820b184c09f98a5a4c884b571963728712591
Author: hubertzhang <hz...@pivotal.io>
Date: 2016-01-27T08:30:13Z
HAWQ-366. Analyse data locality ratio for TPCH query in SFO.
----
> Analyse data locality ratio for TPCH query in SFO.
> --------------------------------------------------
>
> Key: HAWQ-366
> URL: https://issues.apache.org/jira/browse/HAWQ-366
> Project: Apache HAWQ
> Issue Type: Improvement
> Components: Core
> Reporter: Hubert Zhang
> Assignee: Hubert Zhang
>
> Data locality ratio for TPCH query(such as Q2) in SFO is low when some machines are down.
> The reason for this is that we change net_disk_ratio to 1.0 which leads to we prefer the virtual segment with minimum allocated size plus current split size, but not LOCAL!
> To prevent this happens we need to enlarge net_disk_ratio to 1.01. Then if the current size of two vsegs are almost the same, we can choose the LOCAL one.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)